Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!tellab5!chinet!les
From: les@chinet.chi.il.us (Leslie Mikesell)
Newsgroups: comp.sys.att
Subject: Re: Argh! Damn! Kick! (Problem with 3B2 uucp)
Message-ID: <1991Apr24.210028.20581@chinet.chi.il.us>
Date: 24 Apr 91 21:00:28 GMT
References: <1991Apr24.130046.896@cavebbs.gen.nz>
Organization: Chinet - Chicago Public Access UNIX
Lines: 49

In article <1991Apr24.130046.896@cavebbs.gen.nz> clear@cavebbs.gen.nz (Charlie Lear) writes:
>System: 3B2/400, 2 x PORTS cards.

>I tried using the existing modem dialling out on tty12. While the Fido
>program has no hassles, uucp just hangs. Uutry aborts as follows:

[and other modem problems]
If you have SysVr3.1 or later, be sure you are using the M,\M, and \m
fields in your Devices and Dialers files.  Without them, the dialing
will fail unless your modem always holds the Carrier Detect lead up.

Devices entry should look like:
ACU tty12,M - 2400 hayes
Dialers should be something like:
hayes  =,-,  "" \M\pATE1\r\c OK\r \EATDT\T\r\c CONNECT \m\c ""

Note that you can have many entries for the same device using different
speeds, and the speed field can contain a letter to create a "class"
to distinguish it from other entries of the same speed.  With a little
work you can create dummy Systems, Devices, and Dialers entries so that
you can throw modem initialization commands out a port with a simple:
echo ~. |cu dummyname  and it will interlock properly with current
connections.

>Kermit> set line /dev/tty14
>[hang until ctrl-\]
>/dev/tty14: Interrupted System Call

With kermit you have to "set modem" before you "set line" or the same
thing will happen.  If kermit knows how to dial your modem it should
have worked then.

>Someone, somewhere, must surely have had this problem before. I've already
>thrown one modem through the window. I'm debating whether to get the glass
>replaced or if I should throw the 3B2 out through it first. If nobody has 
>any concrete solution I would very much appreciate any pointers on where
>to start looking.

I must say I'm glad I was getting paid for all the time it took me to
get this stuff straightened out....   Maybe I should frame the
AT&T "Closed Problem Report" dated 12/16/86 that explains that they
where adding the sequences to solve a customer's (my) problem.  I had
specifically suggested tokens for the dialer scripts - they threw in
the one for Devices.  I don't know if the choice of the letter had
anything to do with my initial, but at least my boss refers to it
as the "Mikesell" fix.

Les Mikesell
  les@chinet.chi.il.us